A laboratory system for exposing aerosol particles to ozone and rapidly measuring the subsequent changes in their single-particle fluorescence is reported. The system consists of a rotating drum chamber and a single-particle fluorescence spectrometer (SPFS) utilizing excitation at 263 nm. Measurements made with this system show preliminary results on the ultra-violet laser-induced-fluorescence (UV-LIF) spectra of single aerosolized particles of Yersinia rohdei, and of MS2 (bacteriophage) exposed to ozone. When bioparticles are exposed in the chamber the fluorescence emission peak around 330 nm: i) decreases in intensity relative to that of the 400-550 nm band; and ii) shifts slightly toward shorter-wavelengths (consistent with further drying of the particles). In these experiments, changes were observed at exposures below the US Environmental Protection Agency (EPA) limits for ozone
©2012 Optical Society of America
Biological particles in the atmosphere (a subset of organic carbon aerosol) include: bacteria, fungal spores and hyphae, pollens, algae, proteins, viruses, and fragments of the above [1–6]. Bioaerosols may affect the weather and global radiative balance by absorbing and emitting radiation, and can act as cloud-condensation nuclei  and ice-condensation nuclei [6, 8]. Many plants, fungi and microorganisms spread as bioaerosols [e.g., 9]. Some bioaerosols cause infectious diseases and allergies  in humans, other animals and plants. Some bacteria and viruses have been used and/or contemplated as biowarfare agents against humans and their agricultural crops and animals. A variety of methods have been used to study atmospheric bioaerosols. Some techniques used for collections of particles are microscopy; biochemical analyses (e.g., sequencing of DNA or RNA [3, 5]); culturing of bacteria or fungi; mass spectrometry; NMR, IR and Raman spectroscopy ; chromatography; wet-chemical analyses; and fluorescence. Some techniques used for single particles include mass spectrometry; fluorescence; Raman spectroscopy, and elastic scattering. The emphasis in this paper is on the fluorescence properties of biological aerosols, and the effects of aging by ozone in a laboratory setting.
Fluorescence-based bioaerosol measurement systems have been developed by several groups [12–30]. On-line fluorescence measurements of atmospheric aerosols have been made at locations on several continents [4, 14, 15, 19, 20, 26–30]. Many laboratory-based studies of the fluorescence of different bioaerosols have also been performed [17, 23, 31]. Several commercially available systems utilize fluorescence for bioaerosol detection . Interpretation of aerosol fluorescence measurements is difficult, given the many fluorescent molecules that may be in atmospheric particles . Tryptophan is, however, the primary fluorophor responsible for the fluorescence in the 320 to 350 nm range when bacterial bioaerosols are excited by light in the 250 to 290 nm range [12, 33], and tryptophan may be the source of fluorescence in that range for most bioaerosols that contain protein.
There is a large body of literature on atmospheric aging and other atmospheric processing of organic carbon aerosols [e.g . Atmospheric processing of bioaerosols and simulated bioaerosols has been studied using various techniques under several types of study conditions. Originally termed “open air factors”, ozone-olefin mixtures have been shown to be toxic to biological aerosols [35, 36]. Research on the effects of ozone and simulated open air factors (simulated as ozone plus unsaturated terpenes) on aerosolized Micrococcus luteus in a chamber showed that the artificial open air factor was more toxic than either ozone or terpenes alone . Kanaani et al.  studied the changes in fluorescence of fungal spores, measured by a UVAPS, in cultures of different ages, and in cultures exposed to flowing air (prior to aerosolization).
Exposure to ozone affects the viability and inactivation of bacteria on surfaces [39–41]. Ozone treatment can change the absorption and fluorescence of aromatic amino acids (tryptophan, tyrosine, and phenylalanine) in solution [42, 43] and in animal tissues . Ozone treatment has been shown to decrease the tryptophan fluorescence in the 330-nm region . Many of the reports of oxidation of tryptophan by ozone were in aqueous solutions [42, 44–46], often where the ozone was bubbled through the solution. N-formyl kynurinene (NFK), which fluoresces with peak emission near 430 nm, was the primary material formed from oxidation of tryptophan in various studies [42, 44–46]. Ignatenko et al. also found fluorescence-spectral evidence for kynurenine (KU; formed by hydrolysis of NFK) upon oxidation with ozone for longer times .
The effects of ozone on the fluorescence properties of bioaerosols (in air) have not been reported, although measurements of fluorescence spectra of pollens (not aerosolized) treated with ozone have been studied [47, 48]. Attempts to use fluorescence to help determine aerosol composition (e.g., biothreat agents, pollens, fungi, or bacteria, non-biological) are almost exclusively based on the properties of these aerosol shortly after aerosolization. If the fluorescence spectra of aerosols are changed by interaction with pollutants, sunlight, or other environmental factors, the ability to discriminate may be reduced, depending upon the excitation and emission wavelengths, the fluorescence thresholds, and discrimination algorithms used.
Aging of bioaerosols in the natural environment involves many different chemical reactions that occur on multiple time scales (minutes, hours, days) and may involve sequential reactions. To realistically simulate conditions in the natural atmospheric environment for studies of bioaerosol aging, it may be necessary to maintain aerosol populations for long times. Rotating drums have been used since the pioneering work of Goldberg et al.  to maintain aerosol populations entrained in an air mass for minutes to a few days [50–52].
This paper describes a laboratory system developed for studying changes of biological aerosols in simulated atmospheric environments. The system combines: a) a reaction test chamber (rotating drum) to keep particles aloft for significant times (up to 24 hours in some of our measurements); b) an aerosol generator to generate aerosol and inject them into the drum; c) gas generating/modulating equipment, which in the present study generates and controls ozone and water vapor; d) instruments for measuring single-particle fluorescence spectra (similar to Pan et al.  and for measuring integrated fluorescence (TSI’s UVAPS); and e) instruments for measuring ozone concentration, relative humidity and temperature. The paper also describes the use of this system for measuring aerosolized Yersinia rohdei and MS2 bacteriophage, with emphasis on the effects of ozone on the fluorescence properties of these particles. A preliminary description of this system with preliminary results for aerosolized Bacillus thruingiensis was given previously .
Figure 1(a) is a schematic of the overall system showing the aerosol generator; the ozone generator and detector; humidity generator and detector; the laboratory reaction chamber, i.e., the rotating drum (expanded in Fig. 1(b)); the single-particle UV-LIF fluorescence spectrometer (SPFS); and the UVAPS.
2.1 Aerosol generation and control of humidity
Aerosol is generated using an ultrasonic spray nozzle (Sono-Tek Corp., 06–04010), with a 30-mL syringe and pump (Sono-Tek Corp., 11-01061) set to infuse the liquid suspension at a rate of 100 µL/min. The broadband ultrasonic generator (Sono-Tek Corp., 06–05108), used to control the nozzle frequency, is set to 3W. The nozzle of the Sono-Tek is orientated downward into an aerosol capacitance chamber (ACC)  that allows for mixing and initial evaporation of the droplets (initially 18 μm diameter) prior to their entering the drum. HEPA filtered air is then used to carry the generated aerosol from the ACC through a diffusion drier (TSI Inc., 3062) and a Krypton-85 aerosol neutralizer (TSI Inc., 3012). After drying and neutralization, the aerosol laden air is mixed with clean air that has been humidified using a Nafion drier (Perma Pure LLC, PD-50T-24), where, instead of passing dry air outside the nafion membrane, warm saturated air from a heated bubbler is passed around the Nafion bundle in the opposite direction of the flow of gas being humidified . Controlling the ratio of clean humidified air to dry aerosol-laden air is used to produce the sub-saturated RH conditions desired for this study. Although this method limits the maximum RH that can be achieved, it eliminated loss of aerosol to the interior of the Nafion tube bundle, which eliminates the need to clean or replace the Nafion bundles between experiments with different biological material. The final conditioned airstream is introduced to the drum at a flow rate of 10 Lpm.
2.2 Rotating drum
The rotating drum developed and used in this study (Fig. 1(b)) consists of a ~400-L polyvinylchloride (PVC) pipe that is rotated around a fixed center axle. The center axle is composed of clear acrylic and is divided into two concentric sections. The inner axle carries conditioned aerosol laden air into and out of the drum, and the outer axle serves as wire chase so that instrumentation can be fixed to the stationary mount on the axle. The power and data cables can be connected to the exterior without becoming wound around the axle as the drum is rotated. Near the center of the axle, the inner axle opens into the outer axle, where 1/16” holes in the outer axel allow air to move from the outer axel into the main body of the drum. The axle is divided into injection and sampling components by a solid acrylic divide in the center. Conditioned, aerosol-laden air is introduced into the drum via small holes in the injection side of the axle (to the right of the solid divider in Fig. 1). Aged air and aerosol is sampled through a complementary set of small holes on the sampling side of the axle (to the left of the solid divider in Fig. 1). This air can be sampled by a UVAPS (1 Lpm, using the main air flow line and allowing the UVAPS to sample room air for sheath flow), the SPFS (1 Lpm), or by a liquid impinger or filter (10-15 Lpm). Only while aerosol is being generated into the drum, a small fan (not depicted in Fig. 1) is used to mix the aerosol throughout the volume. A HEPA filter connected in series with a vacuum pump was used to evacuate the chamber between experiments.
The interior surfaces of the PVC pipe were sprayed with a Teflon solution in order to provide an inert interior surface. The solvent was allowed to evaporate with one end of the drum open. Any remaining solvent that might interfere with later experiments was removed, after the drum was closed, by reaction for 12 hrs with a high concentration of ozone (>10 ppm) that was generated inside the drum while air was pulled through the drum at 10 Lpm.
The drum can rotate between 1 and 4 rpm/variable. For these experiments the rotation rate was set at approximately 1 rpm in order to maximize the length of time that a 3 µm particle would remain suspended in the drum under ideal conditions .
Despite efforts to minimize the charge on aerosol particles in this study, charge generation by the rotation of the drum was observed to cause the loss of particles during initial experiments. Several additional steps were taken to reduce the amount of charge on the interior surfaces of the drum. A small Polonium source (Amstat Industries Inc., 2U500) was inserted inside the drum on the axis to provide additional ions for neutralization, and grounded copper strips were added inside and outside the drum chamber along the supporting wheel tracks to conduct any charge generated by the turning drum to ground. Electrical current was observed moving through the ground wires using a standard multimeter during drum rotation, indicating that the drum motion was generating a significant amount of charge. It is possible that the electrostatic forces could still have been a factor in the loss of particles, despite the attempts to mitigate the accumulation of charge on the drum surface.
2.3 Ozone generation
Ozone was generated inside the drum partly to minimize the loss of particles that would occur if ozone were pumped into the drum displacing the volume of air in the drum and causing a dilution effect.
A small mercury pen lamp (Pen Ray, 97-0067-01) and an ozone probe (Eco-Sensors, OS-3) were both fixed to the instrument mounting bracket on the axis of the drum. The mercury lamp produces deep ultraviolet light with energy sufficient to photolyze oxygen (wavelengths less than 220 nm). The singlet oxygen formed recombines with molecular oxygen (O2) to form ozone. Because this UV light might interact with the particles directly, causing photochemical reactions, an effort was made to minimize the UV light in the drum and control the amount of ozone produced by covering it with a small metal shield, allowing only a small portion (~3 cm2) of the lamp exposed to the air and was further shielded with a 2 inch diameter Teflon tube that was open at both ends. The ozone generated in this small volume might also be photolyzed by the same UV light to form the highly reactive singlet oxygen O(1D) which can react with water to form the hydroxyl (OH) radical. The role of the hydroxyl radical in these experiments cannot be determined because its production during these experiments was not quantified; however, given limited volume where these interactions could take place, the effect is most likely negligible. No direct measurements of UV light in the drum were made during these experiments, but data previously presented  for Bacillus thruingiensis indicates that the effects of photobleaching, from this light source, on aerosol in the drum is negligible.
2.4 Single-particle fluorescence spectrometer
The single-particle UV-LIF fluorescence spectrometer (SPFS) measurement system was developed and described previously [19, 20]. The system used in this study is the same as the reported SPFS  system but without the concentrator, and with the fluorescence spectrum detector replaced by an Image-Intensified Charge-Coupled Device (ICCD) camera (Andor, Istar DK720-25-UV). Briefly, aerosol is pulled into a 2 × 2–inch optical chamber at 1 L/min from the drum, then focused into a laminar jet of around 300 µm in diameter by a sheath nozzle within the chamber. Any aerosol particles larger than 1 µm flowing through a trigger volume (defined by the intersection of two diode-laser beams) will be detected and illuminated by a single pulse of 263-nm laser (fourth harmonic of Nd:YLF laser, 20 µJ/pulse, 1 mm in diameter; Photonics Industries, DC-150-263). The emitted fluorescence is collected by a Schwarzschild reflective objective (numerical aperture = 0.4) and dispersed by a spectrograph (Acton 150, grating 300 l/mm, blaze wavelength 500 nm). A long-pass filter (cutoff at 280 nm, Newport) in the front of the entrance slit of the spectrograph blocks the elastic scattering from the laser. The dispersed spectrum is recorded by the ICCD camera.
2.5 Ultraviolet Aerodynamic Particle Sizer (UVAPS)
The Ultraviolet Aerodynamic Particle Sizer (UVAPS, TSI, 3314) is commercially available, and has been used by multiple researchers for on-line monitoring of bioaerosols. To help examine the effects of ozone on bioaerosols, a UVAPS is used to measure the aerodynamic particle size distribution and integrated fluorescence (between 430 and 580 nm) intensity distribution excited by a 355-nm laser. Sampling from the drum was limited to 1 Lpm by attaching only the inner, sample-flow inlet to the drum and letting the instrument pull the sheath air (4 Lpm) from the surrounding laboratory.
2.6 Preparation of biological samples
Y. rohdei (ATCC, 43380) incubated at 30 °C for 18 hours in heart infusion broth. After growth the Y. rohdei culture was washed, via centrifugation, 3 times and finally resuspended in 50 mL of phosphate buffered saline (PBS) to produce the final concentration of 9x106 cells/ml. MS2 (ATCC, 15597-B1,) phage was prepared by first incubating E.coli (ATCC, 15597) at 37 °C in EM271 (ATCC) broth. After the culture entered log phase (approximately 3 hours) it was then seeded with MS2 phage. This culture was incubated for an additional 24 hours in EM271 broth at 37 °C. After the second incubation the E. coli cells were pelleted by centrifugation. The supernatant containing the MS2, the E. coli cell lysate, and the spent EM271 culture medium was used for aerosolization. The phage concentration was quantified by plaque assay, at a concentration of 9x1011 pfu/ml, prior to aerosolization. For convenience, these particles will be referred to as MS2 particles throughout the text, even though they contain the E. coli cell lysate and media. These suspensions were loaded into syringes for aerosolization with the ultrasonic nozzle.
2.7 Measurement protocol
Aerosol is continuously introduced into the chamber for 30 min to reach a particle concentration of approximately 1 × 104 per Liter. Air used to carry the aerosol is HEPA filtered and and conditioned using the Nafion humidification system. Once the aerosol concentration has reached at least 1 × 103 particles/L inside the drum, the SPFS and UVAPS are used to measure the properties of aerosol in the drum, while dry, filtered air was allowed into the chamber at 2 L/min to balance the air pulled by the two systems. These measurements are repeated approximately every half hour, with no air being withdrawn from the drum between sampling periods. After the first measurement of particles, ozone is generated until the desired concentration is reached or the end of the experiment. The ozone generator is cycled on and off to maintain the ozone concentration at the desired set point. Measurements for each sample are taken over an approximately 5-hour period. During a few experiments (not reported) spectra from individual aerosol particles were measured even after 24 hours of aging in the rotating drum. Once a sample measurement is finished, and before a new sample is started, the chamber is evacuated at 40 L/min with dry HEPA-filtered air until no particles with diameters above 1 µm are detectable using the UVAPS.
3. Results and discussion
3.1 Characterization of the drum with test aerosols
The drum was initially characterized using Ultrafine Arizona Test Dust (ARD, ISO 12103-1, A2 Powder Techn. Inc.) suspended at a concentration of 10 mg/mL in deionized water. This suspension was aerosolized, as described previously, into the drum and the concentration was measured with a UVAPS over a 12-hr period; 5-min measurements of the concentration were taken with the UVAPS at approximately 30-min intervals (Fig. 2(a) ).
Over the course of the 12-hr period, the number mean diameter (NMD) of the distribution remained relatively constant between 1.03 and 1.30 µm. The geometric standard deviation was between 1.37 and 1.46. After 12 hours in the rotating drum, the aerosol concentration had dropped to 14% of the original concentration. The remaining fraction is more than adequate for making statistically significant measurements of the spectral changes of the aerosol population over time. Part of the decrease in concentration can be attributed to the 25% of the chamber air that was removed for the UVAPS measurements. This does not account for all of the observed loss of particles, but several other factors may have also contributed to the observed loss. There are several deviations in the current drum design from the idealized drum used to develop the equations of Gruel et al. . First, several measurement devices were mounted along the center axis of the drum. These sensors may disrupt the theoretical airflow dynamics in the rotating drum as assumed by Gruel et al.  to cause turbulence, and additional losses. Second, due to the non-conducting nature of the plastic used to construct the drum body and many of the moving parts, a significant amount of static charge was generated inside the drum during rotation, which increased the rate of particle depositional losses due to electrostatic forces.
The biological aerosols of MS2 and Y. rodeii used in this study were mostly in the 1- to 3-µm size range. The number mean diameter, initially between 1.6 and 2.1 µm (depending on the sample), gradually shifted to diameters between 1.2 and 1.6 µm by the end of a typical 5-hour experiment. This apparent shift in the number size distribution is likely caused by the preferential loss of larger particles in the rotating drum, rather than by a change in particle size due to aging. Figure 2(b) shows the number concentration of MS2 particles integrated between 1 and 10 µm. Each value of the aerosol number concentration of MS2 particles is taken as the mean of nine 10-second samples at the specified intervals after the initial fill of the drum. The large loss in concentration between the 0 and 2 hour marks is characteristic of the mixing and initial loss of aerosol in a rotating drum chamber. The rate of concentration decrease varies from sample to sample. Generally, more than 10% of the particles remain suspended after 5 hr. In some samples (not shown), the chamber still had more than 10% of the particles airborne inside of the drum after 24 hr. Figure 2(b) shows the change in particle concentration in the drum over 4 hours. In the first two hours, the concentration decreases by 58%. This is likely the combined effect of particle losses and mixing throughout the chamber volume. Over the next 2 hrs, the concentration dropped by only 20%, which is more representative of the losses alone.
3.2 Fluorescence properties of bioaerosols exposed to ozone
Although efforts were made to clean the drum, the spectra of some particles exhibited a large fluorescence intensity during experiments. Many of these spectra appeared to be consistent with green-fluorescent dye-doped polystyrene spheres that had been used for system calibration. Other particles had fluorescence spectra with huge or saturating intensity but did not appear clearly to be green-fluorescent polystyrene. Both of these types of spectra were omitted manually from the UV-LIF analyses. Such particles likely saturate the UVAPS for the bin in which they fall, but were no more than 1% of the spectra measured.
Figures 3(a)-(c) show typical successive UV-LIF spectra from Y. rohdei and MS2 aerosol particles. Each spectrum is generated from a single particle excited by a single pulse of the 263-nm laser. The largest peak in each spectrum is around 330 nm. The shapes of these spectra are similar to each other. The sharp spikes in each spectrum arise because of read-out and thermal noise of the ICCD, and because the fluorescence from these tiny particles is weak and spread over hundreds of pixels and the amplification required for the ICCD is high. The range of particle sizes (1 to 3 µm in diameter) is responsible for much of the variation in fluorescence intensities for different particles. The fluorescence intensity is proportional to the particle diameter raised to a power that appears to be between 2 and 3 depending upon the wavelengths, particle composition, and size . A diameter raised to a power of 2 has been used effectively for biological particles in the size range used here . Differences in the particle position relative to the collection optics also cause variations in the fraction of fluorescence collected from each particle, and contribute to the variation in the amplitude. Shot-to-shot differences in the laser intensity (less than 10%) also contributed to the variation in fluorescence intensity.
Figures 3(d)-(f) show the averages of 100 single-particle UV-LIF spectra, shown for aerosolized particles of Y. rohdei and MS2 (38% and 43% RH) before and after exposure to ozone for the denoted times. The differences between the ozone-treated and untreated particles are small enough for Y. rohdei and MS2 at 38% RH that without averaging the spectra it is not possible to see clear differences. In these samples, ozone causes a decrease in the 330-nm band relative to the 420-nm band and slightly shifts the 330-nm band to shorter wavelengths. Both the absolute fluorescence intensity and the mean particle size decreased as the measurements progressed. The present measurement system does not provide sufficient data to discern what fraction of the decrease in fluorescence is attributable to changes in particle composition and what fraction results from the decrease in average size of the particles (Figs. 3(d)-3(e)). Most of the spectra are dominated by the tryptophan emission band that peaks near 330 nm. However, in Fig. 3(f) the 420-nm peak is larger than the 330-nm peak for exposure times of 90 minutes or longer.
Differences in the spectral responses of MS2 particles exposed to ozone exposure at different RH values (38% and 43%) are illustrated in Figs. 3(e) and 3(f). In Fig. 3 (f), at an RH of 43%, the fluorescence peak around 330 nm decreased more rapidly than when the RH was 38% (Fig. 3(e)), and shows a smaller decrease for 400-550 nm emission. Similar discernible changes are recognizable in the UV-LIF spectra of individual particles (Fig. 3(b) and (c)). Note, however, in Figs. 3(e) and 3(f) the initial average spectra, prior to ozone exposure, are different: the ratio of the fluorescence-peak-at-330nm (F330) to the fluorescence-peak-near-420nm (F420) is about three in Fig. 3(e), and about two in Fig. 3(f). We do not know the cause of this difference, but note that it makes us less confident in assigning all the differences Figs. 3(e) and 3(f) to the ozone exposure. Also, we only made one run with the higher humidity (43%), and so the data shown is preliminary.
Figure 4 shows the particle size distributions (solid lines) and the averages of the integrated fluorescence from the MS2 particles in each particle size bin (dashed lines) measured by the UVAPS at both 38% and 43% RH before and after 1 hour of ozone exposure. Due to the limited dynamic range (64 bits) of the UVAPS, the detector becomes saturated when the fluorescence intensity count approaches 60, resulting in the flattening of the fluorescence curves for particles in the size range from 2.5 μm to 3.5 μm. In Fig. 4 the average of the integrated fluorescence does not change by more than approximately 10% in any given size bin after 1 hr of exposure. The large decrease in the 430-580-nm fluorescence measured by the SPFS (Fig. 3), cannot be directly compared with these UVAPS results, because the UVAPS data is the fluorescence per particle in a specific size bin, while the SPFS results are average fluorescence spectra for all the particles, for which the mean size decreases, throughout each experiment. Further, the UVAPS excitation at 355 nm does not necessarily excite the same fluorphores as the SPFS (263 nm excitation).
The negligible decrease in the integrated fluorescence measured by the UVAPS is consistent with a hypothesis that the fluorophors emitting light in the 430-580 nm range (when excited by a 355 nm laser) are not significantly changed, in either structure or number, after exposure for one hour. The lack of change in fluorophors responsible for the 355-nm excited fluorescence might be because they are not very susceptible to oxidation by ozone or by reactive species generated in response to ozone under the conditions of this experiment, and/or because they are protected from oxidation (e.g., by being bound to some specific site, such as within an enzyme). One of the primary molecules responsible for fluorescence in the 400- to 550-nm range in many biological particles is NADH [12, 14, 33]. Free NADH is a strong reducing agent with multiple unsaturated bonds. It reacts rapidly with ozone in aqueous solutions . In typical cells the large majority of the NADH is bound to protein (e.g., enzymes used in redox reactions), and the ratio of free NADH to NAD+ is very low, on the order of 0.0015. Bound NADH can still fluoresce, but its susceptibility to oxidation by ozone is not understood. Negligible decrease in the integrated 355 nm fluorescence could also be consistent with changes in fluorophores  (some combination of structure and/or number, possibly including new fluorphors being generated) that did not result in large changes in the total integrated fluorescence between 430 and 580 nm.
The ratio of the fluorescence-peak-at-330nm (F330) to the fluorescence-peak-near-420nm (F420), (i.e., the F330/F420), is a convenient parameter for interpreting the changes in fluorescence amplitudes and spectra as the aerosols are exposed to ozone (Fig. 5 ). In Fig. 5(a) the F330/F420 ratio of Y. rohdei decreases somewhat linearly for the first 300 minutes, and then decreases much more slowly, leaving its F330/F420 at about 50% of its initial value. In Fig. 5(b) the F330/F420 ratio of MS2 decreases to about 75% of its initial value (after 150 minutes of exposure), and then slowly rises back to about 92% of its initial value. In Fig. 5(c) for MS2 at 43% RH, the F330/F420 decreases monotonically, but appears to level off at approximately 25% of its initial value.
During the Y. rohdei experiment, the ozone concentration (Figs. 5(a)-5(c)) increased monotonically from 0 to approximately 9 ppm, over the first 175 minutes of the experiment and was maintained between 8 ppm and 10 ppm over the following 200 minutes. During the MS2 experiment at 38% RH, the ozone increased monotonically from 0 to approximately 10 ppm over the entire 280 minute experiment. During the MS2 experiment at 43% RH, the ozone increased monotonically from 0 to 4 ppm over the 212 minute experiment time.
In all experiments, the F330/F420 ratio decreased relatively rapidly during the initial exposure, and then either decreased more slowly (Figs. 5(d) and 5(f)) or increased slowly (Fig. 5(e)). A possible reason for this rapid-then-slow decline is as follows: the protein structure (primary, secondary and tertiary) can protect some tryptophans from oxidation [59, 60], so that the more-available (or free) tryptophans are oxidized quickly (to products that do not fluoresce at 330 nm), and the more-protected trytophans are not oxidized, or are oxidized slowly, and so keep their 330-nm fluorescence longer.
Further insight into the effects of ozone on fluorescence is obtained by plotting the fluorescence ratio (F330/F420) against the “exposure” (Figs. 5(d)-5(f)), defined as the product of the time of exposure and the average ozone concentration during that time. The US EPA sets an ozone exposure standard of 0.075 ppm, averaged over an 8-hr period. Therefore, the US EPA standard for integrated ozone exposure is 0.075 ppm × 8 hr = 0.6 ppm-hr. In the measurements of MS2 particles at 43% RH (Figs. 3(f), 5(c) and 5(f), the changes in fluorescence spectra occurring with ozone treatment are readily apparent even after the first 30 minutes (the shortest time measured) of exposure to an average ozone concentration of about 0.2 ppm. The concentration-time product for this exposure, i.e., 0.2 ppm × 0.5 hr = 0.1 ppm-hr, is 1/6 the EPA standard for 8 hrs. For this MS2 sample, the decrease in the F330/F420 ratio over this 30-minute period was about 28%. By the time the exposure reached the 0.6 ppm-hr limit (after approximately 80 minutes), the F330/F420 had dropped by about 50%. Over the approximately 35 ppm-hrs of total exposure, the F330/F420 decreased by about 75%.
In these experiments ozone exposures within the US EPA limits result in changes in the spectral shapes of the bioparticles measured. Changes in spectral shape suggest that significant ozone-induced spectral shifts could occur for biological particles in the atmosphere even at ozone concentrations that are well below EPA standards. Typical 8-hr average ozone concentrations in the atmosphere are significantly below the 0.6 ppm-hr standard. Ozone concentrations in pristine air can be as low 0.01 ppm. Because biological particles may remain in the atmosphere for several days, the total exposures of bacteria to ozone could be higher than 0.6 ppm-hrs, even with ozone concentrations well below the 0.075 ppm average for EPA’s 8-hr limit. For example, if the average concentration were 0.02 ppm for 3 days, the total exposure would be 1.44 ppm-hrs.
Aqueous tryptophan can be converted by ozone to NFK and KU, both of which fluoresce in the 400-420 nm range ; however, in the present experiments there does not seem to be an increase in fluorescence due to NFK and KU. This lack of an increase in fluorescence might be due to the lack of available water in these bioparticles at both of the RH used here (38% and 43%). The potential mechanism follows Prior and Uppa . A secondary ozonide is formed in the reaction of ozone with the pyrrole ring of tryptophan. Hydrolysis of this ozonide to form the fluorescent molecules NFK, KU, etc., requires water (e.g., Fig. 1 in Pryor and Uppa ). Studies of the oxidation of tryptophan at low RH have not been performed. However, if the mechanism of Pryor and Uppa  applies, then at sufficiently low RH, the oxidation with ozone should stop at the secondary ozonide. Although the fluorescence of the suggested secondary ozonide product has not been measured, its structure indicates that it has less conjugation than either tryptophan or KU, and so is unlikely to fluoresce with peak emission longer than 330 nm.
The cause of the change in spectra with only small changes in RH for MS2 particles does not have a straightforward explanation. We only made one run with the higher RH (43%), and the initial average spectra, prior to ozone exposure, are different, and so we do not want to over-interpret the data. However, diffusion of ozone into and/or through the particle may be faster at higher relative humidity, consistent with the observations of Shiraiwa et al.  of a more rapid uptake of ozone by amorphous protein particles at higher humidity. Also, the cascade mechanism of Pryor et al. [63,64] might be relevant, as it depends upon water to generate hydrogen peroxide. Pryor et al.  showed that reactive oxygen species, such as peroxides or reactive aldehydes, are generated when O3 reacts with olefins (e.g., in cell membranes). This cascade mechanism can generate one mole of hydrogen peroxide per mole of ozone, olefin and water (see Eq. (4) in Pryor et al. ). The hydrogen peroxide generated can continue the cascade and oxidize other molecules. Although not discussed by Pryor, at sufficiently low humidity, the rate of hydrogen peroxide production would be limited by water, and so the cascade would be stopped there. Whether that applies to the fluorescence measurements in this paper is unknown.
A laboratory system was developed for the measurement of changes in the fluorescence of biological particles in simulated atmospheric environments. Samples of Y. rohdei and MS2-in-E.coli lysate were aerosolized, exposed to ozone in a rotating drum, and their fluorescence was measured using 263-nm and 355-nm excitation. The results indicate that the fluorescence emission peak around 330 nm appears to decrease in intensity and become slightly blue-shifted. The fluorescence peak in the 400- to 550-nm range decreases less rapidly than the 330 nm tryptophan peak. There are distinguishable changes in the UV-LIF spectra of single MS2 particles that were exposed to ozone. Exposure of MS2 particles to ozone at a slightly higher humidity resulted in more rapid decreases in the fluorescence both near 330 nm and in the 400–550 nm range. The 330-nm band was observed to decrease more rapidly than the 400- to 550-nm band at the higher RH. No significant changes were seen in the integrated fluorescence (430 nm to 580 nm) measured by the UVAPS (excitation wavelength of 355 nm; Fig. 4). These experiments have led to several observations about the oxidation of biological particles; however, the exact mechanisms that lead to the observed changes in spectra are not immediately obvious.
These changes in fluorescence may significantly impact fluorescence-based detection of biological aerosols that were prepared in a similar manner and exposed to ozone for comparable concentration-time products. The impact may apply even when multiple fluorescence emission bands are used to identify the aerosol particles.
This research was supported by the Defense Threat Reduction Agency under contract number HDTRA1-10-C-0023, and by US Army Research Laboratory mission funds. We thank Dr. Sari Paikoff (DTRA) for her support. We thank Susan Wu and Jerry Hahn for growing the bacteria used in this study.
References and links
1. B. Lighthart and A. J. Mohr, eds., Atmospheric Microbial Aerosols (Chapman & Hall, 1994).
2. W. Elbert, P. E. Taylor, M. O. Andreae, and U. Poschl, “Contribution of fungi to primary biogenic aerosols in the atmosphere: wet and dry discharged spores, carbohydrates, and inorganic ions,” Atmos. Chem. Phys. 7(17), 4569–4588 (2007). [CrossRef]
3. J. Fröhlich-Nowoisky, D. A. Pickersgill, V. R. Després, and U. Pöschl, “High diversity of fungi in air particulate matter,” Proc. Natl. Acad. Sci. U.S.A. 106(31), 12814–12819 (2009). [CrossRef] [PubMed]
4. J. A. Huffman, B. Treutlein, and U. Poschl, “Fluorescent biological aerosol particle concentrations and size distributions measured with an Ultraviolet Aerodynamic Particle Sizer (UV-APS) in Central Europe,” Atmos. Chem. Phys. 10(7), 3215–3233 (2010). [CrossRef]
5. R. M. Bowers, A. P. Sullivan, E. K. Costello, J. L. Collett Jr, R. Knight, and N. Fierer, “Sources of bacteria in outdoor air across cities in the midwestern united states,” Appl. Environ. Microbiol. 77(18), 6350–6356 (2011). [CrossRef] [PubMed]
6. V. Despres, A. J. Huffman, S. M. Burrows, C. Hoose, A. S. Safatov, G. Buryak, J. Frohlich-Nowoisky, W. Elbert, M. O. Andraea, U. Poschl, and R. Jaenicke, “Primary biological aerosol particles in the atmosphere: a review,” Tellus B Chem. Phys. Meterol. 64(15598), 1–58 (2012).
7. J. M. Sun and P. A. Ariya, “Atmospheric organic and bio-aerosols as cloud condensation nuclei (CCN): A review,” Atmos. Environ. 40(5), 795–820 (2006). [CrossRef]
8. R. M. Bowers, C. L. Lauber, C. Wiedinmyer, M. Hamady, A. G. Hallar, R. Fall, R. Knight, and N. Fierer, “Characterization of airborne microbial communities at a high-elevation site and their potential to act as atmospheric ice nuclei,” Appl. Environ. Microbiol. 75(15), 5121–5130 (2009). [CrossRef] [PubMed]
9. J. M. Prospero, E. Blades, G. Mathison, and R. Naidu, “Interhemispheric transport of viable fungi and bacteria from Africa to the Caribbean with soil dust,” Aerobiologia 21(1), 1–19 (2005). [CrossRef]
10. T. Ichinose, S. Yoshida, K. Hiyoshi, K. Sadakane, H. Takano, M. Nishikawa, I. Mori, R. Yanagisawa, H. Kawazato, A. Yasuda, and T. Shibamoto, “The effects of microbial materials adhered to Asian sand dust on allergic lung inflammation,” Arch. Environ. Contam. Toxicol. 55(3), 348–357 (2008). [CrossRef] [PubMed]
11. A. Tripathi, R. E. Jabbour, J. A. Guicheteau, S. D. Christesen, D. K. Emge, A. W. Fountain, J. R. Bottiger, E. D. Emmons, and A. P. Snyder, “Bioaerosol analysis with raman chemical imaging microspectroscopy,” Anal. Chem. 81(16), 6981–6990 (2009). [CrossRef] [PubMed]
12. S. C. Hill, R. G. Pinnick, P. Nachman, G. Chen, R. K. Chang, M. W. Mayo, and G. L. Fernandez, “Aerosol-fluorescence spectrum analyzer: real-time measurement of emission spectra of airborne biological particles,” Appl. Opt. 34(30), 7149–7155 (1995). [CrossRef] [PubMed]
13. P. P. Hairston, J. Ho, and F. R. Quant, “Design of an instrument for real-time detection of bioaerosols using simultaneous measurement of particle aerodynamic size and intrinsic fluorescence,” J. Aerosol Sci. 28(3), 471–482 (1997). [CrossRef] [PubMed]
14. J. Ho, and Y-W., December 23, 1997 (Filed March 19, 1996). Fluorescent biological particle detection system. United States patent 5,701,012.
15. M. Seaver, J. D. Eversole, J. J. Hardgrove, W. K. Cary Jr, and D. C. Roselle, “Size and Fluorescence Measurements for Field Detection of Biological Aerosols,” Aerosol Sci. Technol. 30(2), 174–185 (1999). [CrossRef]
16. P. H. Kaye, W. R. Stanley, E. Hirst, E. V. Foot, K. L. Baxter, and S. J. Barrington, “Single particle multichannel bio-aerosol fluorescence sensor,” Opt. Express 13(10), 3583–3593 (2005). [CrossRef] [PubMed]
17. A. Manninen, M. Putkiranta, J. Saarela, A. Rostedt, T. Sorvajärvi, J. Toivonen, M. Marjamäki, J. Keskinen, and R. Hernberg, “Fluorescence cross sections of bioaerosols and suspended biological agents,” Appl. Opt. 48(22), 4320–4328 (2009). [CrossRef] [PubMed]
18. Y. L. Pan, J. Hartings, R. G. Pinnick, S. C. Hill, J. Halverson, and R. K. Chang, “Single-particle fluorescence spectrometer for ambient aerosols,” Aerosol Sci. Technol. 37(8), 628–639 (2003). [CrossRef]
19. Y. L. Pan, R. G. Pinnick, S. C. Hill, and R. K. Chang, “Particle-fluorescence spectrometer for real-time single-particle measurements of atmospheric organic carbon and biological aerosol,” Environ. Sci. Technol. 43(2), 429–434 (2009). [CrossRef] [PubMed]
20. Y. L. Pan, S. C. Hill, R. G. Pinnick, H. Huang, J. R. Bottiger, and R. K. Chang, “Fluorescence spectra of atmospheric aerosol particles measured using one or two excitation wavelengths: Comparison of classification schemes employing different emission and scattering results,” Opt. Express 18(12), 12436–12457 (2010). [CrossRef] [PubMed]
21. F. L. Reyes, T. H. Jeys, N. R. Newbury, C. A. Primmerman, G. S. Rowe, and A. Sanchez, “Bio-aerosol fluorescence sensor,” Field Anal. Chem. Technol. 3(4-5), 240–248 (1999). [CrossRef]
23. V. Sivaprakasam, H.-B. Lin, A. L. Huston, and J. D. Eversole, “Spectral characterization of biological aerosol particles using two-wavelength excited laser-induced fluorescence and elastic scattering measurements,” Opt. Express 19(7), 6191–6208 (2011). [CrossRef] [PubMed]
24. D. P. Greenwood, T. H. Jeys, B. Johnson, J. M. Richardson, and M. P. Shatz, “Optical Techniques for Detecting and Identifying Biological-Warfare Agents,” Proc. IEEE 97(6), 971–989 (2009). [CrossRef]
25. K. Mitsumoto, K. Yabusaki, K. Kobayashi, and H. Aoyagi, “Development of a novel real-time pollen-sorting counter using species-specific pollen autofluorescence,” Aerobiologia 26(2), 99–111 (2010). [CrossRef]
26. A. Birenzvige, J. Eversole, M. Seaver, S. Francesconi, E. Valdes, and H. Kulaga, “Aerosol Characteristics in a Subway Environment,” Aerosol Sci. Technol. 37(3), 210–220 (2003). [CrossRef]
27. R. G. Pinnick, S. C. Hill, Y. L. Pan, and R. K. Chang, “Fluorescence spectra of atmospheric aerosol at Adelphi, Maryland, USA; measurement and classification of single particles containing organic carbon,” Atmos. Environ. 38(11), 1657–1672 (2004). [CrossRef]
28. Y.-L. Pan, R. G. Pinnick, S. C. Hill, J. M. Rosen, and R. K. Chang, “Single-particle laser-induced-fluorescence spectra of biological and other organic-carbon aerosols in the atmosphere: Measurements at New Haven, Connecticut, and Las Cruces, New Mexico,” J. Geophys. Res. 112(D24), D24S19 (2007). [CrossRef]
29. A. M. Gabey, M. W. Gallagher, J. Whitehead, J. Dorsey, P. H. Kaye, and W. R. Stanley, “Measurements and comparison of primary biological aerosol above and below a tropical forest canopy using a dual-channel fluorescence aerosol spectrometer,” Atmos. Chem. Phys. 10(10), 4453–4466 (2010). [CrossRef]
30. A. M. Gabey, W. R. Stanley, M. W. Gallagher, and P. H. Kaye, “The fluorescence properties of aerosol larger than 0.8μm in urban and tropical rainforest locations,” Atmos. Chem. Phys. 11(11), 5491–5504 (2011). [CrossRef]
31. V. Agranovski, Z. Ristovski, M. Hargreaves, P. J. Blackall, and L. Morawska, “Performance evaluation of the UVAPS: influence of physiological age of airborne bacteria and bacterial stress,” J. Aerosol Sci. 34(12), 1711–1727 (2003). [CrossRef]
32. R. DeFreez, “LIF bio-aerosol threat triggers: then and now,” Proc. SPIE 7484, 74840H (15 pp.)(2009).
33. C. Pöhlker, J. A. Huffman, and U. Pöschl, “Autofluorescence of atmospheric bioaerosols – fluorescent biomolecules and potential interferences,” Atmos. Meas. Tech. Discuss. 4(5), 5857–5933 (2011). [CrossRef]
35. W. D. Sawyer, J. V. Jemski, A. L. Hogge Jr, H. T. Eigelsbach, E. K. Wolfe, H. G. Dangerfield, W. S. Gochenour Jr, and D. Crozier, “Effect of Aerosol Age on the Infectivity of Airborne Pasteurella tularensis for Macaca mulatta and Man,” J. Bacteriol. 91(6), 2180–2184 (1966). [PubMed]
37. R. Bailey, L. Fielding, A. Young, and C. Griffith, “Effect of Ozone and Open Air Factor against Aerosolized Micrococcus luteus,” J. Food Prot. 70(12), 2769–2773 (2007). [PubMed]
38. H. Kanaani, M. Hargreaves, Z. Ristovski, and L. Morawska, “Performance assessment of UVAPS: Influence of fungal spore age and air exposure,” J. Aerosol Sci. 38(1), 83–96 (2007). [CrossRef]
40. J. G. Kim and A. E. Yousef, “Inactivation kinetics of foodborne spoilage and pathogenic bacteria by ozone,” J. Food Sci. 65(3), 521–528 (2000). [CrossRef]
41. M. V. Selma, A. M. Ibáñez, M. Cantwell, and T. Suslow, “Reduction by gaseous ozone of salmonella and microbial flora associated with fresh-cut cantaloupe,” Food Microbiol. 25(4), 558–565 (2008). [CrossRef] [PubMed]
42. A. V. Ignatenko, B. A. Tatarinov, N. N. Khovratovich, V. P. Khrapovitskii, and S. N. Cherenkevich, “Spectral-fluorescent investigation of the action of ozone on aromatic amino acids,” J. Appl. Spectrosc. 37(1), 781–784 (1982).
43. A. V. Ignatenko, “Use of the method of tryptophan fluorescence to characterize disruptions of the structure of ozonized proteins,” J. Appl. Spectrosc. 49(1), 691–695 (1988). [CrossRef]
44. E. Fujimori, “Changes induced by ozone and ultraviolet light in type I collagen. Bovine Achilles tendon collagen versus rat tail tendon collagen,” Eur. J. Biochem. 152(2), 299–306 (1985). [CrossRef] [PubMed]
46. T. Kotiaho, M. N. Eberlin, P. Vainiotalo, and R. Kostiainen, “Electrospray mass and tandem mass spectrometry identification of ozone oxidation products of amino acids and small peptides,” J. Am. Soc. Mass Spectrom. 11(6), 526–535 (2000). [CrossRef] [PubMed]
47. V. V. Roshchina and E. V. Melnikova, “Pollen chemosensitivity to ozone and peroxides,” Russ. J. Plant Physiol. 48(1), 74–83 (2001). [CrossRef]
48. V. V. Roshchina, “Autofluorescence of plant secreting cells as a biosensor and bioindicator reaction,” J. Fluoresc. 13(5), 403–420 (2003). [CrossRef]
49. L. J. Goldberg, H. M. S. Watkins, E. E. Boerke, and M. A. Chatigny, “The use of a rotating drum for the study of aerosols over extended periods of time,” Am. J. Hyg. 68(1), 85–93 (1958). [PubMed]
50. V. Krumins, E.-K. Son, G. Mainelis, and D. E. Fennell, “Retention of Inactivated Bioaerosols and Ethene in a Rotating Gioreactor Constructed for Bioaerosol Activity Studies,” Clean 36(7), 593–600 (2008).
52. Y. L. Pan, S. C. Hill, R. G. Pinnick, J. L. Santarpia, N. Baker, B. Alvarez, S. Ratnesar-Shumate, B. Cottrell, and L. McKee, “Fluorescence spectra of bioaerosols exposed to ozone in a laboratory reaction chamber to simulate atmospheric processing,” Proc. SPIE 8018, 801804, 801804-7 (2011). [CrossRef]
53. S. R. Ratnesar-Shumate, M. L. Wagner, C. Kerechanin, G. House, K. M. Brinkley, C. Bare, N. Baker, R. Quizon, J. Quizon, A. Proescher, E. Van Gieson, and J. L. Santarpia, “Improved method for the evaluation of real-time biological aerosol detection technologies,” Aerosol Sci. Technol. 45(5), 635–644 (2011). [CrossRef]
54. J. L. Santarpia, R. Gasparini, R. Li, and D. R. Collins, “Diurnal variations in the hygroscopic growth cycles of ambient aerosol populations,” J. Geophys. Res. 110(D3), D03206 (2005). [CrossRef]
55. R. L. Gruel, C. R. Reid, and R. T. Allemann, “The optimum rate of drum rotation for aerosol aging,” J. Aerosol Sci. 18(1), 17–22 (1987). [CrossRef]
56. S. C. Hill, R. G. Pinnick, S. Niles, N. F. Fell, Y. L. Pan, J. Bottiger, B. V. Bronk, S. Holler, and R. K. Chang, “Fluorescence from airborne microparticles: dependence on size, concentration of fluorophores, and illumination intensity,” Appl. Opt. 40(18), 3005–3013 (2001). [CrossRef] [PubMed]
58. Y. L. Pan, R. G. Pinnick, S. C. Hill, S. Niles, S. Holler, J. R. Bottiger, J.-P. Wolf, and R. K. Chang, “Dynamics of photon-induced degradation and fluorescence in riboflavin microparticles,” Appl. Phys. B 72(4), 449–454 (2001). [CrossRef]
59. J. B. Mudd, P. J. Dawson, S. Tseng, and F. P. Liu, “Reaction of ozone with protein tryptophans: band III, serum albumin, and cytochrome C,” Arch. Biochem. Biophys. 338(2), 143–149 (1997). [CrossRef] [PubMed]
60. B. S. Berlett, R. L. Levine, and E. R. Stadtman, “Comparison of the effects of ozone on the modification of amino acid residues in glutamine synthetase and bovine serum albumin,” J. Biol. Chem. 271(8), 4177–4182 (1996). [CrossRef] [PubMed]
61. M. Shiraiwa, M. Ammann, T. Koop, and U. Pöschl, “Gas uptake and chemical aging of semisolid organic aerosol particles,” Proc. Natl. Acad. Sci. U.S.A. 108(27), 11003–11008 (2011). [CrossRef] [PubMed]
62. W. A. Pryor and R. M. Uppu, “A kinetic model for the competitive reactions of ozone with amino acid residues in proteins in reverse micelles,” J. Biol. Chem. 268(5), 3120–3126 (1993). [PubMed]
64. W. A. Pryor, G. L. Squadrito, and M. Friedman, “The cascade mechanism to explain ozone toxicity: the role of lipid ozonation products,” Free Radic. Biol. Med. 19(6), 935–941 (1995). [CrossRef] [PubMed]